Reclaiming Local Data:
Building Community-Led AI Projects
Why Local Data Matters
AI development is often framed as the domain of big tech, driven by global platforms and massive datasets. But for many communities, the most valuable insights aren’t hidden in the world’s largest corporation, they’re found closer to home.
From local transport patterns to oral histories, local data is the lifeblood of projects that aim to reflect community needs and identities. When communities and SMEs take ownership of their data, they can build AI systems that are relevant, transparent, and trusted.
The Pitfalls of Global Platforms
Relying exclusively on global AI providers introduces risks:
Generic outputs that miss local nuance
Data extraction that strips value away from communities
Opaque processes that make accountability difficult
For sectors like heritage, education, and civic engagement, these gaps can undermine trust and relevance.
Building Community-Led AI
Grassroots organisations and SMEs are well placed to take a different path. Some practical starting points:
Data Co-Creation: Collecting and curating datasets with community participation
Small-scale models: Training lighter, domain-specific systems on local infrastructure
Shared ownership: Establishing governance models where data and outcomes remain in community hands
Towards Local Resilience
Community-led AI won’t replace global platforms, but it can rebalance them. By reclaiming local data, communities gain not only better tools but also greater agency over the technologies shaping their lives.
